In the current study, hypotheses regarding the structure of social support and the role of the structure of social support in the stressor-illness model will be examined. A two-wave panel survey of a representative sample (N=500) of adults (18-65) residing in a SMSA in upstate New York will be conducted. The structure of social support is viewed to consist of two components: (1) resources and (2) functions. Two levels of resources are postulated to exist: interpersonal ties and community level ties. Two types of functions are postulated to operate: expressive and instrumental. The structure of social support is identified by the relationship between its two basis components, with resources causally preceding and influencing functions (resources---greater than functions). Four different social support processes are specified: actual (received) routine, actual (received) crisis, perceived routine, and perceived crisis. Once the structure of social support is identified, it will be modelled in the context of life events and health (mental, physical, and general well being) using single indicator (correlational, regression, path analysis) and multiple indicator (structural equation modelling with LISREL) analysis. Control variables include psychological resources and key demographic factors. Mediating, counteracting, and interacting effects of support will be examined in the context of the stressor illness model. It is postulated that the heretofore weak or non existing mediating effect of one of the components of social support (resources) is due to fact that the majority of its effect on mental and physical health is indirect (via the function component). The proposed study will require four years of work.